Hi, thanks for releasing amazing work!
I have a question about the sigma schedule, which is different between the code and paper.
In the paper, sigma is defined as $\frac{t'}{t_{thresh}}$ in Eq. (11), which is also mentioned in another issue. However, in code, it seems sigma is only $t'$, which is different to paper.
I am wondering if there is some misunderstanding? If the code is correct, does it mean the same sigma schedule is also used during training?